A time-delay neural network architecture for isolated word recognition

By Kevin J. Lang, A. Waibel, Geoffrey E. Hinton. 1990

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Deep learning in neural networks: An overview (2014, background): “Some of the GPUbased FNN implementations (Sections 5.16–5.19) have greatly contributed to recent successes in contests for pattern recognition (Sections 5.19–5.22), image segmentation (Section 5.21), and object detection (Sections 5.21–5.22).”
Deep Learning: Methods and Applications (2014, background): “It is useful to point out that the time-delay neural network (TDNN) [202, 382] developed for early speech recognition is a special case and predecessor of the CNN when weight sharing is limited to one of the two dimensions, i.”
A tutorial survey of architectures, algorithms, and applications for deep learning (2014, background): “It is useful to point out that time-delay neural networks (TDNN, [127, 129]) developed for early speech recognition are a special case of the CNNwhenweight sharing is limited to one of the twodimensions, i.”
Speech Recognition via Large-Scale Weak Supervision (unknown): “Second were a number of ways of handling warp-H et al. 1989, Lang et al. 1990), RNNs (Robinson and Fallside, 1991), and the hybrid hybrid HMM/MLP architecture in which a feedforward neural network is trained as a phonetic classiﬁer whose outputs are used as probability estimates for an HMM-based…”
Evolving Artiicial Neural Networks (2007, background): “There have been some successful applications of BP in various areas [20, 21, 22], but BP has drawbacks due to its use of gradient descent [23, 24].”
Drug Design by Machine Learning: Support Vector Machines for Pharmaceutical Data Analysis (2001, background): “Neural networks have been successfully applied in many settings, including speech recognition (Lang et al., 1990), handwritten digit recognition (LeCun et al.”
Evolving artificial neural networks (1999, background): “There have been some successful applications of BP in various areas [20]–[22], but BP has drawbacks due to its use of gradient descent [23], [24].”
Optimizing Neural Networks that Generate Iimages (2014, background): “In audio recognition, most systems use a lot of domain-specific knowledge: the Fourier transform, Cepstral transform, and sometimes convolution over the time (Lang et al., 1990) or frequency domains (Abdel-Hamid et al.”
An analysis of particle swarm optimizers (2002, background): “Past applications include, amongst others, the recognition of handwritten characters [77], learning to recognise spoken words [75] and recognising images of faces [25].”
DeepCare: A Deep Dynamic Memory Model for Predictive Medicine (2016, methodology): “Since the ﬁrst version of RNN was introduced in the 1980s [33], many varieties of RNN have been proposed such as Time-Delay Neural Networks [34] and Echo State Network [35].”